The switch supports fault resistance by allowing a standby switch to take over if the active switch becomes unavailable. Cisco nonstop forwarding (NSF) works with stateful switchover (SSO) to minimize the amount of time a network is unavailable.

NSF provides these benefits:

• Improved network availability—NSF continues forwarding network traffic and application state information so that user session information is maintained after a switchover.

• Overall network stability—Network stability may be improved with the reduction in the number of route flaps, which were created when routers in the network failed and lost their routing tables.

• Neighboring routers do not detect a link flap—Because the interfaces remain up during a switchover, neighboring routers do not detect a link flap (the link does not go down and come back up).

• Prevents routing flaps—Because SSO continues forwarding network traffic during a switchover, routing flaps are avoided.

• Maintains user sessions established prior to the switchover.

When a standby switch runs in SSO mode, the standby switch starts up in a fully-initialized state and synchronizes with the persistent configuration and the running configuration of the active switch. It subsequently maintains the state on the protocols listed below, and all changes in hardware and software states for features that support stateful switchover are kept in synchronization. Consequently, it offers minimum interruption to Layer 2 sessions in a redundant active switch configuration.

If the active switch fails, the standby switch becomes the active switch. This new active switch uses existing Layer 2 switching information to continue forwarding traffic. Layer 3 forwarding will be delayed until the routing tables have been repopulated in the newly active switch.

Note	SSO is not supported if the IOS-XE software is running the LAN Base license level.

The state of these features is preserved between both the active and standby switches:

• 802.3
• 802.3u
• 802.3x (Flow Control)
• 802.3ab (GE)
• 802.3z (Gigabit Ethernet including CWDM)
• 802.3ad (LACP)
• 802.1p (Layer 2 QoS)
• 802.1q
• 802.1X (Authentication)
• 802.1D (Spanning Tree Protocol)
• 802.3af (Inline power)
• PAgP
• VTP
• Dynamic ARP Inspection
• DHCP snooping
• IP source guard
• IGMP snooping (versions 1 and 2)
• DTP (802.1q and ISL)
• MST
• PVST+
• Rapid-PVST
• PortFast/UplinkFast/BackboneFast
• BPDU guard and filtering
• Voice VLAN
• Port security
• Unicast MAC filtering
• ACL (VACLS, PACLS, RACLS)
• QOS (DBL)
• Multicast storm control/broadcast storm control

SSO is compatible with the following list of features. However, the protocol database for these features is not synchronized between the standby and active switches:

• 802.1Q tunneling with Layer 2 Protocol Tunneling (L2PT)
• Baby giants
• Jumbo frame support
• CDP
• Flood blocking
• UDLD
• SPAN/RSPAN
• NetFlow

All Layer 3 protocols on a switch are learned on the standby switch if SSO is enabled.

Cisco IOS Nonstop Forwarding (NSF) always runs with stateful switchover (SSO) and provides redundancy for Layer 3 traffic. NSF is supported by the BGP, OSPF, and EIGRP routing protocols and is supported by Cisco Express Forwarding (CEF) for forwarding. The routing protocols have been enhanced with NSF-capability and awareness, which means that routers running these protocols can detect a switchover and take the necessary actions to continue forwarding network traffic and to recover route information from the peer devices.

Each protocol depends on CEF to continue forwarding packets during switchover while the routing protocols rebuild the Routing Information Base (RIB) tables. After the routing protocols have converged, CEF updates the FIB table and removes stale route entries. CEF then updates the hardware with the new FIB information.

If the active switch is configured for BGP (with the graceful-restart command), OSPF, or EIGRP routing protocols, routing updates are automatically sent during the active switch election.

The switch supports NSF-awareness and NSF-capability for the BGP, OSPF, and EIGRP protocols in IP Services license level and NSF-awareness for the EIGRP-stub in IP Base license level.

NSF has two primary components:

• NSF-awareness

  A networking device is NSF-aware if it is running NSF-compatible software. If neighboring router devices detect that an NSF router can still forward packets when an active switch election happens, this capability is referred to as NSF-awareness. Cisco IOS enhancements to the Layer 3 routing protocols (BGP, OSPF, and EIGRP) are designed to prevent route-flapping so that the CEF routing table does not time out or the NSF router does not drop routes. An NSF-aware router helps to send routing protocol information to the neighboring NSF router. NSF-awareness is enabled by default for EIGRP-stub, EIGRP, and OSPF protocols. NSF-awareness is disabled by default for BGP.

• NSF-capability

  A device is NSF-capable if it has been configured to support NSF; it rebuilds routing information from NSF-aware or NSF-capable neighbors. NSF works with SSO to minimize the amount of time that a Layer 3 network is unavailable following an active switch election by continuing to forward IP packets. Reconvergence of Layer 3 routing protocols (BGP, OSPFv2, and EIGRP) is transparent to the user and happens automatically in the background. The routing protocols recover routing information from neighbor devices and rebuild the Cisco Express Forwarding (CEF) table.

  Note	NSF does not support IPv6 and is IPv4 Unicast only.

A key element of Cisco IOS Nonstop Forwarding (NSF) is packet forwarding. In a Cisco networking device, packet forwarding is provided by Cisco Express Forwarding (CEF). CEF maintains the FIB and uses the FIB information that was current at the time of the switchover to continue forwarding packets during a switchover. This feature reduces traffic interruption during the switchover.

During normal NSF operation, CEF on the active supervisor switch synchronizes its current FIB and adjacency databases with the FIB and adjacency databases on the standby switch. Upon switchover, the standby switch initially has FIB and adjacency databases that are mirror images of those that were current on the active switch. CEF keeps the forwarding engine on the standby switch current with changes that are sent to it by CEF on the active switch. The forwarding engine can continue forwarding after a switchover as soon as the interfaces and a data path are available.

As the routing protocols start to repopulate the RIB on a prefix-by-prefix basis, the updates cause prefix-by-prefix updates to CEF, which it uses to update the FIB and adjacency databases. Existing and new entries receive the new version (“epoch”) number, indicating that they have been refreshed. The forwarding information is updated on the forwarding engine during convergence. The switch signals when the RIB has converged. The software removes all FIB and adjacency entries that have an epoch older than the current switchover epoch. The FIB now represents the newest routing protocol forwarding information.

When an NSF-capable router begins a BGP session with a BGP peer, it sends an OPEN message to the peer. Included in the message is a statement that the NSF-capable device has “graceful” restart capability. Graceful restart is the mechanism by which BGP routing peers avoid a routing flap following a switchover. If the BGP peer has received this capability, it is aware that the device sending the message is NSF-capable. Both the NSF-capable router and its BGP peers need to exchange the graceful restart capability in their OPEN messages at the time of session establishment. If both the peers do not exchange the graceful restart capability, the session will not be capable of a graceful restart.

If the BGP session is lost during the active switch switchover, the NSF-aware BGP peer marks all the routes associated with the NSF-capable router as stale; however, it continues to use these routes to make forwarding decisions for a set period of time. This functionality prevents packets from being lost while the newly active switch is waiting for convergence of the routing information with the BGP peers.

After an active switch switchover occurs, the NSF-capable router reestablishes the session with the BGP peer. In establishing the new session, it sends a new graceful restart message that identifies the NSF-capable router as having restarted.

At this point, the routing information is exchanged between the two BGP peers. After this exchange is complete, the NSF-capable device uses the routing information to update the RIB and the FIB with the new forwarding information. The NSF-aware device uses the network information to remove stale routes from its BGP table; the BGP protocol then is fully converged.

If a BGP peer does not support the graceful restart capability, it ignores the graceful restart capability in an OPEN message but establishes a BGP session with the NSF-capable device. This function allows interoperability with non-NSF-aware BGP peers (and without NSF functionality), but the BGP session with non-NSF-aware BGP peers is not capable of a graceful restart.

Note

BGP support in NSF requires that neighbor networking devices be NSF-aware; that is, the devices must have the graceful restart capability and advertise that capability in their OPEN message during session establishment. If an NSF-capable router discovers that a particular BGP neighbor does not have graceful restart capability, it does not establish an NSF-capable session with that neighbor. All other neighbors that have graceful restart capability continue to have NSF-capable sessions with this NSF-capable networking device.

When an OSPF NSF-capable router performs an active switch switchover, it must perform the following tasks in order to resynchronize its link state database with its OSPF neighbors:

• Relearn the available OSPF neighbors on the network without causing a reset of the neighbor relationship
• Reacquire the contents of the link state database for the network

As quickly as possible after an active switch switchover, the NSF-capable router sends an OSPF NSF signal to neighboring NSF-aware devices. Neighbor networking devices recognize this signal as an indicator that the neighbor relationship with this router should not be reset. As the NSF-capable router receives signals from other routers on the network, it can begin to rebuild its neighbor list.

After neighbor relationships are reestablished, the NSF-capable router begins to resynchronize its database with all of its NSF-aware neighbors. At this point, the routing information is exchanged between the OSPF neighbors. Once this exchange is complete, the NSF-capable device uses the routing information to remove stale routes, update the RIB, and update the FIB with the new forwarding information. The OSPF protocols are then fully converged.

Note

OSPF support in NSF requires that all neighbor networking devices be NSF-aware. If an NSF-capable router discovers that it has non-NSF -aware neighbors on a particular network segment, it disables NSF capabilities for that segment. Other network segments composed entirely of NSF-capable or NSF-aware routers continue to provide NSF capabilities.

When an EIGRP NSF-capable router initially re-boots after an NSF restart, it has no neighbor and its topology table is empty. The router is notified by the standby (now active) switch when it needs to bring up the interfaces, reacquire neighbors, and rebuild the topology and routing tables. The restarting router and its peers must accomplish these tasks without interrupting the data traffic directed toward the restarting router. EIGRP peer routers maintain the routes learned from the restarting router and continue forwarding traffic through the NSF restart process.

To prevent an adjacency reset by the neighbors, the restarting router uses a new Restart (RS) bit in the EIGRP packet header to indicate a restart. The RS bit is set in the hello packets and in the initial INIT update packets during the NSF restart period. The RS bit in the hello packets allows the neighbors to be quickly notified of the NSF restart. Without seeing the RS bit, the neighbor can only detect an adjacency reset by receiving an INIT update or by the expiration of the hello hold timer. Without the RS bit, a neighbor does not know if the adjacency reset should be handled using NSF or the normal startup method.

When the neighbor receives the restart indication, either by receiving the hello packet or the INIT packet, it recognizes the restarting peer in its peer list and maintains the adjacency with the restarting router. The neighbor then sends it topology table to the restarting router with the RS bit set in the first update packet indicating that it is NSF-aware and is helping out the restarting router. The neighbor does not set the RS bit in their hello packets, unless it is also a NSF restarting neighbor.

Note	A router may be NSF-aware but may not be helping the NSF restarting neighbor because booting from a cold start.

If at least one of the peer routers is NSF-aware, the restarting router would then receive updates and rebuild its database. The restarting router must then find out if it had converged so that it can notify the routing information base (RIB). Each NSF-aware router is required to send an end of table (EOT) marker in the last update packet to indicate the end of the table content. The restarting router knows it has converged when it receives the EOT marker. The restarting router can then begin sending updates.

An NSF-aware peer would know when the restarting router had converged when it receives an EOT indication from the restarting router. The peer then scans its topology table to search for the routes with the restarted neighbor as the source. The peer compares the route timestamp with the restart event timestamp to determine if the route is still available. The peer then goes active to find alternate paths for the routes that are no longer available through the restarted router.

When the restarting router has received all EOT indications from its neighbors or when the NSF converge timer expires, EIGRP notifies the RIB of convergence. EIGRP waits for the RIB convergence signal and then floods its topology table to all awaiting NSF-aware peers.